Application of Speech Technology in the Multilingual Speedata Project